Distribution Grid Power Flow Calculation Using Segmented Backward-Forward Sweep

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current power flow calculation methods for distribution grids, such as the backward/forward sweep algorithm, are inefficient and resource-intensive, especially when dealing with renewable energy sources like photovoltaic panels, due to high computational time and sensitivity to grid characteristics like R/X ratios, making them impractical for complex distribution systems.

Innovation Solution

The method involves dividing the distribution grid into middle voltage and low voltage grids, pre-processing low voltage grids to estimate total equivalent power, and performing a backward/forward sweep algorithm only on the middle voltage grid, refining low voltage grid parameters based on middle voltage results, which includes summing loads, sources, and line losses, and iteratively calculating node voltages and branch currents until convergence is reached.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If backward/forward sweep algorithm is applied to calculate power flow in distribution grid, then power flow calculation can be performed, but computational time and computer resources increase significantly when dealing with complex grids containing over 5000 components and renewable energy sources

Engineering Contradiction:
Improvepower flow calculation accuracyVSAvoidcomputational time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The distribution grid is divided into multiple feeders, and each feeder is processed independently. This segmentation reduces the computational complexity from handling the entire grid at once to processing smaller, manageable sections, thereby reducing computational time and resource requirements while maintaining calculation accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The algorithm performs a backward sweep first to calculate total power at departure nodes before performing the forward sweep to calculate voltages. This preliminary action of calculating power flows first enables more efficient subsequent voltage calculations and allows for early termination conditions to be evaluated.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If backward/forward sweep algorithm processes all components in the network including photovoltaic panels, then complete power flow analysis is achieved, but the method becomes impractical due to high computational complexity and resource requirements

Engineering Contradiction:
Improvepower flow analysis completenessVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

Photovoltaic panels and other distributed energy resources are extracted and handled separately from the main backward/forward sweep algorithm. Their power injections are calculated independently based on their operational characteristics, then incorporated into the power flow calculation. This extraction reduces the complexity of the main algorithm while maintaining complete analysis of all grid components.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The algorithm performs power flow calculation only on the necessary portions of the grid at each iteration level, rather than recalculating all components every iteration. By calculating power flows feeder-by-feeder and only updating affected sections, the method achieves complete analysis with reduced computational complexity.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If backward/forward sweep algorithm is used for power flow calculation, then results such as node voltage, branch current and line loss can be obtained, but the convergence becomes sensitive to grid characteristics such as R/X ratios in distribution systems

Engineering Contradiction:
Improvepower flow calculation resultsVSAvoidalgorithm convergence
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The algorithm applies different calculation approaches to different parts of the grid based on local characteristics. Feeders with high R/X ratios use adjusted calculation methods, while feeders with low R/X ratios use standard methods. This local adaptation improves convergence reliability across diverse grid conditions while maintaining calculation precision.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The algorithm iteratively calculates power flows and voltages, using the results of each iteration to improve the next. The backward sweep calculates power flows based on current voltage estimates, then the forward sweep updates voltage estimates based on calculated power flows. This feedback loop continues until convergence criteria are met, ensuring reliable convergence even with varying R/X ratios.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3098924B1A method and device for calculation of power flow in the distribution grid
Publication Date: 2020.10.28 ELECTRICITE DE FRANCE
  • EP3098924B1 patent drawingFigure 1
  • EP3098924B1 patent drawingFigure 2
  • EP3098924B1 patent drawingFigure 3a

AI summary

The invention relates to calculation of power flow in a distribution grid, comprising the steps of: - dividing (S1*) the distribution grid into two parts, middle voltage grid and one or more low voltage grids; - pre-processing (S2*) all low voltage grids in order to estimate a total equivalent power; - performing (S3*) a backward/forward type sweep algorithm upon the middle voltage grid only; - after middle voltage power flow computation (S4*), refining (S5*) all low voltage grid parameters based on middle voltage grid parameter obtained in the previous step, said parameter being one of voltage, current, power, and impedance.